CN109100562B - Voltage flicker parameter detection method based on complex value independent component analysis - Google Patents
Voltage flicker parameter detection method based on complex value independent component analysis Download PDFInfo
- Publication number
- CN109100562B CN109100562B CN201810969807.4A CN201810969807A CN109100562B CN 109100562 B CN109100562 B CN 109100562B CN 201810969807 A CN201810969807 A CN 201810969807A CN 109100562 B CN109100562 B CN 109100562B
- Authority
- CN
- China
- Prior art keywords
- formula
- voltage flicker
- signal
- component analysis
- independent component
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/12—Measuring rate of change
Abstract
The invention relates to a voltage flicker parameter detection method based on a plurality of independent component analysis, which is characterized by comprising the following steps: the method comprises the steps of establishing a complex value independent component analysis model of the voltage flicker signal, eliminating the uncertainty of amplitude and phase in the complex value independent component analysis model, and calculating the voltage flicker parameter in the complex value independent component analysis model. Has the advantages of scientific and reasonable structure, simple structure, high precision, wide application range and the like.
Description
Technical Field
The invention belongs to the technical field of electric energy quality parameter detection, and relates to a voltage flicker parameter detection method based on complex value independent component analysis.
Background
The voltage flicker reflects the influence of unstable (flickering) light brightness caused by voltage fluctuation on visual sensation of human eyes, and according to literature data statistics, the frequency range which can be perceived by people on the voltage flicker of the incandescent lamp is 1-25 Hz, the sensitive frequency range is 6-12 Hz, and the human eyes are most sensitive to the voltage flicker with the frequency near 8.8 Hz. The influence of voltage flicker on the production and life of people is mainly reflected in that: the flicker of lighting light can be caused, so that the vision of people is easy to fatigue, and the vision is influenced; the brightness, color change and image shaking of the picture displayed by the electronic display equipment can be caused, and the picture display quality is reduced; the rotating speed of the motor is not uniform, the product quality is influenced, the service life of equipment is prolonged, the reliable operation level is reduced, and the production cost of enterprises is increased; in addition, the normal operation of equipment errors such as computers, automatic control equipment, electronic instruments and the like can be influenced, even serious derived accidents are caused, and the serious influence is caused on the life of people and the production of enterprises. In order to reduce the influence of voltage flicker on an electric power system and improve the electric energy quality, firstly, the voltage flicker signal parameters need to be measured to realize the evaluation of the voltage flicker, and in addition, the distortion of the electric energy signal needs to be reduced by adopting a corresponding method according to the parameters of the voltage flicker signal, so the estimation of the voltage flicker parameters is particularly important and is a precondition for improving the electric energy quality.
The voltage flicker parameter detection and analysis method is a popular research direction both at home and abroad, and various new algorithms are continuously emerged along with the development of mathematical theories:
1) the voltage flicker parameter detection method based on Fourier transform is the most common voltage flicker parameter detection method, accurate frequency and amplitude parameters of voltage flicker signals can be obtained by performing fast Fourier transform on the voltage flicker signals subjected to synchronous sampling, asynchronous sampling can be caused when frequency fluctuation of flicker components of a power grid is large, and under the condition of the asynchronous sampling, frequency spectrum leakage and fence phenomena can be generated by the Fourier transform, so that the voltage flicker parameter estimation error is large. In order to solve the problem of Fourier transform in asynchronous sampling, people propose that the problem of reducing frequency spectrum leakage by adopting interpolation, a Hamming window, a Blackman window and a Gaussian window function is solved;
2) the voltage flicker parameter detection method based on the spatial spectrum estimation mainly adopts array signal processing methods such as a rotation invariant signal parameter estimation method (ESPRIT), a Prony algorithm (Rrony) and a multiple signal classification (MUSIC) which are commonly used in the spectrum estimation to realize the estimation of the voltage flicker parameter, the method has higher frequency resolution and error performance, a characteristic value decomposition and root searching process is required to be carried out, and when the frequency of a signal to be estimated is more, the calculation amount is obviously increased;
3) a voltage flicker parameter detection method based on wavelet transformation. The wavelet transform has good localization characteristics and is called as a mathematical microscope, but different wavelet functions are overlapped in a frequency domain, so that the error is large when the frequency detection of a voltage flicker signal is carried out, the performance of the algorithm is greatly influenced by the selection of the wavelet basis functions, and the most appropriate wavelet basis function is difficult to select under the condition of power grid signal fluctuation;
besides the voltage flicker parameter detection method, methods based on S transformation, Hilbert-Huang transformation, Chip-Z transformation, Teager energy operator, real independent component analysis and the like are also available. The existing methods are generally only effective for voltage flicker with unchanged parameters, and the methods for estimating dynamic voltage flicker parameters are less, for example: methods based on a space spectrum estimation method, Fourier transform and the like are only suitable for estimating static parameters of voltage flicker, most algorithms cannot directly give voltage flicker parameters, and parameter information needs to be obtained through methods such as spectrum peak search and the like. The voltage flicker detection method based on real number independent component analysis needs to separate flicker envelopes firstly, then utilizes Fourier transform to analyze the envelopes to obtain parameter information, has a complex process, and can not be used for the condition of dynamic change of voltage flicker parameters.
Disclosure of Invention
The invention aims to provide a voltage flicker parameter detection method based on complex value independent component analysis, which has the advantages of simple structure, high precision and wide application range, aiming at the problems that the voltage flicker parameter detection method has a complex structure and low precision and is not suitable for the condition of dynamic parameter change.
The purpose of the invention is realized by the following technical scheme: a voltage flicker secondary parameter detection method based on complex value independent component analysis is characterized by comprising the following steps: establishing a complex value independent component analysis model of the voltage flicker signal, eliminating the uncertainty of amplitude and phase in the complex value independent component analysis model, and calculating the voltage flicker parameter in the complex value independent component analysis model:
1) establishment of complex value independent component analysis model of voltage flicker signal
The method comprises the following steps of expanding a single-path voltage flicker signal into multiple paths of signals by utilizing a time delay and filtering method, enabling the signals to meet the requirement of a complex value independent component analysis model on the number of signals to be processed, enabling the voltage flicker parameter model to be a multi-input single-output model, enabling the complex value independent component analysis signal processing model to be a multi-input multi-output model, and in order to separate each component of the flicker signal by adopting the complex value independent component analysis, unifying the voltage flicker parameter model and the complex value independent component analysis signal processing model, and mathematically expressing the instantaneous voltage containing the voltage flicker into a form of an amplitude modulation signal:
in the formula, A0Is the amplitude value of the 50Hz power frequency voltage, n is the number of flicker components, AiIs the amplitude, ω, of the ith voltage flicker componentiIs the i-th voltage flicker component angular frequency, t is the time variable, θiIs the phase of the i-th voltage flicker component, ω0Is the angular frequency, theta, of the mains voltage0The phase of the power frequency voltage is expanded to obtain a voltage flicker signal expression:
further finishing the formula (2) to obtain:
in the formula, D0＝A0，
When i ≠ 0:
vi＝ω0+ωi，
εi＝ω0-ωi，φi＝θ0-θithe frequencies of the voltage flicker components are all less than 50Hz, so that v in the formula (3)iFor frequency components greater than 50Hz,. epsiloniFor frequency components less than 50Hz, viIs a frequency component greater than 50Hz and epsiloniFor frequency components less than 50Hz, each containing parameter information for an equal amount of voltage flicker component, vi＝ω0+ωiAnd εi＝ω0-ωiBoth are related, resulting in
And Dicos(εit+φi) The two terms are also related, in the course of model building,
and Dicos(εit+φi) The two terms are respectively used for constructing the source signal, if not
And Dicos(εit+φi) One of the two terms is filtered, correlation exists between the constructed information source signals, the requirement that complex value independent component analysis needs to meet mutual statistical independence between the information source signals is not met, therefore, a filter is used for filtering frequency components larger than 50Hz or frequency components smaller than 50Hz, and high-pass filtering is used for filtering epsilon in the formula (3)iI.e. frequency components less than 50Hz, while retaining v in the formula (3)iI.e. frequency components greater than 50Hz, the filtered voltage signal is expressed as:
performing Hilbert transform on the formula (4) to obtain a complex expression of the formula:
the complex independent component analysis model is a complex value multiple input multiple output model, and the formula (5) can be regarded as a multiple input single output model, multiple inputs refer to voltage flicker components and power frequency signals, single output refers to only 1 channel of observation signals, in order to estimate voltage flicker parameters from the formula (5) by utilizing the complex independent component analysis, a plurality of virtual observation signals are constructed by carrying out equal delay processing on the formula (5), namely:
in the formula, delta t is delay time, and the above formula is further finished to obtain
In the formula (I), the compound is shown in the specification,
Ci＝exp(jviΔ t), then m delays are performed to obtain:
in the formula, Cim＝exp(jvimΔt)，j2And (3) constructing a plurality of time delay signals by using the formula (8), and further constructing a complex independent component analysis model for detecting the voltage flicker parameters, namely:
X(t)＝AS(t) (9)
wherein x (t) ═ x (t), x (t + Δ t), …, x (t + m Δ t)]T，A＝[a0,a1…,am]T，ai＝[P0C0i,P1C1i,…,PnCni]，，C00＝C10＝…Cn0＝1，S(t)＝[exp(jvot),…,exp(jvnt)]TThe purpose of detecting the voltage flicker parameter is to estimate the components of the voltage flicker signal from the observed signal X (t), i.e. to estimate Ai、ωi，
2) Elimination of amplitude and phase uncertainty in complex-valued independent component analysis models
And (3) constructing a constraint condition by using prior information in the complex value independent component analysis model, eliminating uncertainty in the newly-built model through the constraint condition, and decomposing the mixing matrix A in the formula (9) to obtain:
if in formula (9)
Then the expression of equation (9) is:
in the formula, si(t)＝Piexp(jvit), by observing the formula (12), it can be found that: the 1 st row of the hybrid matrix is 1, so that the condition is taken as a constraint condition, the estimation matrix of the hybrid matrix is compensated, each column of the estimation matrix of the hybrid matrix is divided by the element of the corresponding first row, the 1 st row of the estimation matrix of the hybrid matrix is forced to meet the condition, and the uncertainty of the amplitude and the phase of the estimation matrix of the hybrid matrix is eliminated;
assuming that the voltage flicker parameter is estimated by adopting the analysis of the complex independent components, the directly obtained estimation matrix of the hybrid matrix is V, and the expression is as follows:
in the formula, βNIs any real number, N is 0,1, …, N, βNCausing uncertainty in the amplitude of the complex hybrid matrix, thetaNIs an arbitrary phase, θNAnd (3) causing uncertainty of the phase of the complex mixing matrix, and decomposing the formula (13) to obtain:
(14) the right term 1 of the equation equal sign is the mixing matrix a in equation (12), so equation (14) is written as:
in the formula (I), the compound is shown in the specification,
the mixing matrix a can be expressed as:
for matrix V0The first row of the estimation matrix of the mixing matrix obtained primarily by the complex independent component analysis processing is transformed into V0The specific implementation method is that a diagonal matrix with the size consistent with the column number of the mixed matrix is constructed, the elements of the diagonal matrix consist of the first row elements of the (13) type matrix V, and the estimated mixed matrix A is consistent with the real mixed matrix A, so that each component of the estimated voltage flicker is consistent with the real voltage flicker component;
3) calculation of voltage flicker parameters in complex-valued independent component analysis models
Calculating voltage flicker parameters from the estimated mixing matrix and the source signal, and separating the voltage flicker signals by adopting complex value independent component analysis on the basis of a complex value independent component analysis model of the voltage flicker signals to obtain a mixing matrix A in the complex value independent component analysis model, namely a formula (12) in which C isim＝exp(jvim Δ t), the mixing matrix is expressed as:
from the above formula, the ratio between two adjacent rows of elements in any column is:
the corresponding angular frequency is thus determined:
in the formula, QIrAnd QRrAre respectively
Imaginary and real parts, averaged for improved stability, i.e.
According to vi＝w0+wi，w0The frequency is the frequency of a 50Hz power frequency signal, the frequencies obtained by the formula (20) are sequenced, and the frequency obtained by subtracting the power frequency from other frequencies except the minimum frequency is the frequency of the voltage flicker component;
in terms of amplitude estimation, the expression of the source signal in the complex-valued independent component analysis model can be found from equation (12):
si(t)＝Piexp(jvit) (21)
in the formula (I), the compound is shown in the specification,
D0＝A0，A0is the amplitude of the power frequency signal, AiThat is, the amplitude of the voltage flicker signal component, and thus the amplitude of the flicker signal is obtained by taking the modulus value of the estimated signal obtained by analysis of the complex-valued independent component, i.e., the amplitude of the voltage flicker signal component
Di＝|si(t)|＝|Piexp(jvit)| (22)
Because the amplitude of the power frequency signal is far greater than that of the flicker signal component, D obtained by the formula (22) is obtainediSorting is carried out, the maximum value is the amplitude of the power frequency signal, DiAnd multiplying the other values except the maximum value by 2 times after sorting to obtain the amplitude of the voltage flicker signal.
The invention relates to a voltage flicker parameter detection method based on complex value independent component analysis, which comprises the following steps of: firstly, carrying out high-pass filtering on the voltage flicker signal, and filtering out low-frequency signal components with the frequency less than 50 HZ; then, an equal-interval delay method is sampled to construct a multi-path virtual observation signal, and the multi-path virtual observation signal is converted into a complex signal by performing Hilbert transform on the multi-path virtual observation signal; the multi-channel complex signals correspond to mixed signals in the complex value independent component analysis model, each component signal of each channel of complex signals corresponds to an information source signal in the complex value independent component analysis model, and a linear matrix between each component signal and the multi-channel signals corresponds to a mixed matrix in the complex value independent component analysis model; in terms of model uncertainty elimination: the prior information of a newly-built model is utilized to convert a mixed matrix in the model into a mixed matrix with all elements of a 1 st row being 1, and the constraint condition is used as a characteristic constraint condition of the mixed matrix to eliminate uncertainty in the model; in terms of parameter estimation: calculating frequency information of voltage flicker by using the estimated ratio relation between two adjacent rows in the same column of the hybrid matrix; and calculating the amplitude information of the signal by using a method for calculating and estimating the modulus value of the signal. Has the advantages of scientific and reasonable structure, simple structure, high precision, wide application range and the like.
Drawings
FIG. 1 is a flow chart of a voltage flicker parameter detection method based on complex-valued independent component analysis.
Detailed Description
The invention is further illustrated by the following figures and detailed description.
Referring to fig. 1, the invention relates to a voltage flicker secondary parameter detection method based on complex value independent component analysis, which comprises the following steps: establishing a complex value independent component analysis model of the voltage flicker signal, eliminating the uncertainty of amplitude and phase in the complex value independent component analysis model, and calculating the voltage flicker parameter in the complex value independent component analysis model:
1) establishment of complex value independent component analysis model of voltage flicker signal
The method comprises the following steps of expanding a single-path voltage flicker signal into multiple paths of signals by utilizing a time delay and filtering method, enabling the signals to meet the requirement of a complex value independent component analysis model on the number of signals to be processed, enabling the voltage flicker parameter model to be a multi-input single-output model, enabling the complex value independent component analysis signal processing model to be a multi-input multi-output model, and in order to separate each component of the flicker signal by adopting the complex value independent component analysis, unifying the voltage flicker parameter model and the complex value independent component analysis signal processing model, and mathematically expressing the instantaneous voltage containing the voltage flicker into a form of an amplitude modulation signal:
in the formula, A0Is the amplitude value of the 50Hz power frequency voltage, n is the number of flicker components, AiIs the amplitude, ω, of the ith voltage flicker componentiIs the i-th voltage flicker component angular frequency, t is the time variable, θiIs the phase of the i-th voltage flicker component, ω0Is the angular frequency, theta, of the mains voltage0The phase of the power frequency voltage is expanded to obtain a voltage flicker signal expression:
further finishing the formula (2) to obtain:
in the formula, D0＝A0，v0＝ω0，
When i ≠ 0:
vi＝ω0+ωi，
εi＝ω0-ωi，φi＝θ0-θivoltage flash componentThe frequencies of the quantities are all less than 50Hz, so that v in the formula (3)iFor frequency components greater than 50Hz,. epsiloniFor frequency components less than 50Hz, viIs a frequency component greater than 50Hz and epsiloniFor frequency components less than 50Hz, each containing parameter information for an equal amount of voltage flicker component, vi＝ω0+ωiAnd εi＝ω0-ωiBoth are related, resulting in
And Dicos(εit+φi) The two terms are also related, in the course of model building,
and Dicos(εit+φi) The two terms are respectively used for constructing the source signal, if not
And Dicos(εit+φi) One of the two terms is filtered, correlation exists between the constructed information source signals, the requirement that complex value independent component analysis needs to meet mutual statistical independence between the information source signals is not met, therefore, a filter is used for filtering frequency components larger than 50Hz or frequency components smaller than 50Hz, and high-pass filtering is used for filtering epsilon in the formula (3)iI.e. frequency components less than 50Hz, while retaining v in the formula (3)iI.e. frequency components greater than 50Hz, the filtered voltage signal is expressed as:
performing Hilbert transform on the formula (4) to obtain a complex expression of the formula:
the complex independent component analysis model is a complex value multiple input multiple output model, and the formula (5) can be regarded as a multiple input single output model, multiple inputs refer to voltage flicker components and power frequency signals, single output refers to only 1 channel of observation signals, in order to estimate voltage flicker parameters from the formula (5) by utilizing the complex independent component analysis, a plurality of virtual observation signals are constructed by carrying out equal delay processing on the formula (5), namely:
in the formula, delta t is delay time, and the above formula is further finished to obtain
In the formula (I), the compound is shown in the specification,
Ci＝exp(jviΔ t), then m delays are performed to obtain:
in the formula, Cim＝exp(jvimΔt)，j2And (3) constructing a plurality of time delay signals by using the formula (8), and further constructing a complex independent component analysis model for detecting the voltage flicker parameters, namely:
X(t)＝AS(t) (9)
wherein x (t) ═ x (t), x (t + Δ t), …, x (t + m Δ t)]T，A＝[a0,a1…,am]T，ai＝[P0C0i,P1C1i,…,PnCni]，，C00＝C10＝…Cn0＝1，S(t)＝[exp(jvot),…,exp(jvnt)]TThe purpose of detecting the voltage flicker parameter is to estimate the components of the voltage flicker signal from the observed signal X (t), i.e. to estimate Ai、ωi，
2) Elimination of amplitude and phase uncertainty in complex-valued independent component analysis models
And (3) constructing a constraint condition by using prior information in the complex value independent component analysis model, eliminating uncertainty in the newly-built model through the constraint condition, and decomposing the mixing matrix A in the formula (9) to obtain:
if in formula (9)
Then the expression of equation (9) is:
in the formula, si(t)＝Piexp(jvit), by observing the formula (12), it can be found that: the 1 st row of the hybrid matrix is 1, so that the condition is taken as a constraint condition, the estimation matrix of the hybrid matrix is compensated, each column of the estimation matrix of the hybrid matrix is divided by the element of the corresponding first row, the 1 st row of the estimation matrix of the hybrid matrix is forced to meet the condition, and the uncertainty of the amplitude and the phase of the estimation matrix of the hybrid matrix is eliminated;
assuming that the voltage flicker parameter is estimated by adopting the analysis of the complex independent components, the directly obtained estimation matrix of the hybrid matrix is V, and the expression is as follows:
in the formula, βNIs any real number, N is 0,1, …, N, βNCausing uncertainty in the amplitude of the complex hybrid matrix, thetaNIs an arbitrary phase, θNAnd (3) causing uncertainty of the phase of the complex mixing matrix, and decomposing the formula (13) to obtain:
(14) the right term 1 of the equation equal sign is the mixing matrix a in equation (12), so equation (14) is written as:
in the formula (I), the compound is shown in the specification,
the mixing matrix a can be expressed as:
for matrix V0The first row of the estimation matrix of the mixing matrix obtained primarily by the complex independent component analysis processing is transformed into V0The specific implementation method is that a diagonal matrix with the size consistent with the column number of the mixed matrix is constructed, the elements of the diagonal matrix consist of the first row elements of the (13) type matrix V, and the estimated mixed matrix A is consistent with the real mixed matrix A, so that each component of the estimated voltage flicker is consistent with the real voltage flicker component;
3) calculation of voltage flicker parameters in complex-valued independent component analysis models
Calculating voltage flicker parameters from the estimated mixing matrix and the source signal, and separating the voltage flicker signals by adopting complex value independent component analysis on the basis of a complex value independent component analysis model of the voltage flicker signals to obtain a mixing matrix A in the complex value independent component analysis model, namely a formula (12) in which C isim＝exp(jvim Δ t), the mixing matrix is expressed as:
from the above formula, the ratio between two adjacent rows of elements in any column is:
the corresponding angular frequency is thus determined:
in the formula, QIrAnd QRrAre respectively
Imaginary and real parts, averaged for improved stability, i.e.
According to vi＝w0+wi，w0The frequency is the frequency of a 50Hz power frequency signal, the frequencies obtained by the formula (20) are sequenced, and the frequency obtained by subtracting the power frequency from other frequencies except the minimum frequency is the frequency of the voltage flicker component;
in terms of amplitude estimation, the expression of the source signal in the complex-valued independent component analysis model can be found from equation (12):
si(t)＝Piexp(jvit) (21)
in the formula (I), the compound is shown in the specification,
D0＝A0，A0is the amplitude of the power frequency signal, AiThat is, the amplitude of the voltage flicker signal component, and thus the amplitude of the flicker signal is obtained by taking the modulus value of the estimated signal obtained by analysis of the complex-valued independent component, i.e., the amplitude of the voltage flicker signal component
Di＝|si(t)|＝|Piexp(jvit)| (22)
Because the amplitude of the power frequency signal is far greater than that of the flicker signal component, D obtained by the formula (22) is obtainediSorting is carried out, the maximum value is the amplitude of the power frequency signal, DiMultiplying the values except the maximum value by 2 after sortingThe multiple is the amplitude of the voltage flicker signal.
The software routines of the present invention are programmed according to automation, networking and computer processing techniques, and are well known to those skilled in the art.
The embodiments of the present invention are provided for illustration only and not for the purpose of limitation, and it should be understood by those skilled in the art that various modifications and equivalent changes made by reference to the embodiments of the present invention are within the scope of the claims.
Claims (1)
1. A voltage flicker parameter detection method based on complex value independent component analysis is characterized by comprising the following steps: establishing a complex value independent component analysis model of the voltage flicker signal, eliminating the uncertainty of amplitude and phase in the complex value independent component analysis model, and calculating the voltage flicker parameter in the complex value independent component analysis model:
1) establishment of complex value independent component analysis model of voltage flicker signal
The method comprises the following steps of expanding a single-path voltage flicker signal into multiple paths of signals by utilizing a time delay and filtering method, enabling the signals to meet the requirement of a complex value independent component analysis model on the number of signals to be processed, enabling the voltage flicker parameter model to be a multi-input single-output model, enabling the complex value independent component analysis signal processing model to be a multi-input multi-output model, and in order to separate each component of the flicker signal by adopting the complex value independent component analysis, unifying the voltage flicker parameter model and the complex value independent component analysis signal processing model, and mathematically expressing the instantaneous voltage containing the voltage flicker into a form of an amplitude modulation signal:
in the formula, A0Is the amplitude value of the 50Hz power frequency voltage, n is the number of flicker components, AiIs the amplitude, ω, of the ith voltage flicker componentiIs the i-th voltage flicker component angular frequency, t is the time variable, θiIs the phase of the i-th voltage flicker component，ω0Is the angular frequency, theta, of the mains voltage0The phase of the power frequency voltage is expanded to obtain a voltage flicker signal expression:
further finishing the formula (2) to obtain:
in the formula, D0＝A0，v0＝ω0，
When i ≠ 0:
vi＝ω0+ωi，
εi＝ω0-ωi，φi＝θ0-θithe frequencies of the voltage flicker components are all less than 50Hz, so that v in the formula (3)iFor frequency components greater than 50Hz,. epsiloniFor frequency components less than 50Hz, viIs a frequency component greater than 50Hz and epsiloniFor frequency components less than 50Hz, each containing parameter information for an equal amount of voltage flicker component, vi＝ω0+ωiAnd εi＝ω0-ωiBoth are related, resulting in
And Dicos(εit+φi) The two terms are also related, in the course of model building,
and Dicos(εit+φi) The two terms are respectively used for constructing the source signal, if not
And Dicos(εit+φi) One of the two terms is filtered, correlation exists between the constructed information source signals, the requirement that complex value independent component analysis needs to meet mutual statistical independence between the information source signals is not met, therefore, a filter is used for filtering frequency components larger than 50Hz or frequency components smaller than 50Hz, and high-pass filtering is used for filtering epsilon in the formula (3)iI.e. frequency components less than 50Hz, while retaining v in the formula (3)iI.e. frequency components greater than 50Hz, the filtered voltage signal is expressed as:
performing Hilbert transform on the formula (4) to obtain a complex expression of the formula:
the complex independent component analysis model is a complex value multiple input multiple output model, and the formula (5) can be regarded as a multiple input single output model, multiple inputs refer to voltage flicker components and power frequency signals, single output refers to only 1 channel of observation signals, in order to estimate voltage flicker parameters from the formula (5) by utilizing the complex independent component analysis, a plurality of virtual observation signals are constructed by carrying out equal delay processing on the formula (5), namely:
in the formula, delta t is delay time, and the above formula is further finished to obtain
In the formula (I), the compound is shown in the specification,
Ci＝exp(jviΔ t), then m delays are performed to obtain:
in the formula, Cim＝exp(jvimΔt)，j2And (3) constructing a plurality of time delay signals by using the formula (8), and further constructing a complex independent component analysis model for detecting the voltage flicker parameters, namely:
X(t)＝AS(t) (9)
wherein x (t) ═ x (t), x (t + Δ t), …, x (t + m Δ t)]T，A＝[a0,a1…,am]T，ai＝[P0C0i,P1C1i,…,PnCni]，C00＝C10＝…Cn0＝1，S(t)＝[exp(jvot),…,exp(jvnt)]TThe purpose of detecting the voltage flicker parameter is to estimate the components of the voltage flicker signal from the observed signal X (t), i.e. to estimate Ai、ωi，
2) Elimination of amplitude and phase uncertainty in complex-valued independent component analysis models
And (3) constructing a constraint condition by using prior information in the complex value independent component analysis model, eliminating uncertainty in the newly-built model through the constraint condition, and decomposing the mixing matrix A in the formula (9) to obtain:
if in formula (9)
Then the expression of equation (9) is:
in the formula, si(t)＝Piexp(jvit), by observing the formula (12), it can be found that: the 1 st row of the hybrid matrix is 1, so that the condition is taken as a constraint condition, the estimation matrix of the hybrid matrix is compensated, each column of the estimation matrix of the hybrid matrix is divided by the element of the corresponding first row, the 1 st row of the estimation matrix of the hybrid matrix is forced to meet the condition, and the uncertainty of the amplitude and the phase of the estimation matrix of the hybrid matrix is eliminated;
assuming that the voltage flicker parameter is estimated by adopting the analysis of the complex independent components, the directly obtained estimation matrix of the hybrid matrix is V, and the expression is as follows:
in the formula, βNIs any real number, N is 0,1, …, N, βNCausing uncertainty in the amplitude of the complex hybrid matrix, thetaNIs an arbitrary phase, θNAnd (3) causing uncertainty of the phase of the complex mixing matrix, and decomposing the formula (13) to obtain:
(14) the right term 1 of the equation equal sign is the mixing matrix a in equation (12), so equation (14) is written as:
in the formula (I), the compound is shown in the specification,
the mixing matrix a can be expressed as:
for matrix V0The first row of the estimation matrix of the mixing matrix obtained primarily by the complex independent component analysis processing is transformed into V0The specific implementation method is that a diagonal matrix with the size consistent with the column number of the mixed matrix is constructed, the elements of the diagonal matrix consist of the first row elements of the (13) type matrix V, and the estimated mixed matrix A is consistent with the real mixed matrix A, so that each component of the estimated voltage flicker is consistent with the real voltage flicker component;
3) calculation of voltage flicker parameters in complex-valued independent component analysis models
Calculating voltage flicker parameters from the estimated mixing matrix and the source signal, and separating the voltage flicker signals by adopting complex value independent component analysis on the basis of a complex value independent component analysis model of the voltage flicker signals to obtain a mixing matrix A in the complex value independent component analysis model, namely a formula (12) in which C isim＝exp(jvim Δ t), the mixing matrix is expressed as:
from the above formula, the ratio between two adjacent rows of elements in any column is:
the corresponding angular frequency is thus determined:
in the formula, QIrAnd QRrAre respectively
Imaginary and real parts ofStability is improved by taking the average value thereof, i.e.
According to vi＝w0+wi，w0The frequency is the frequency of a 50Hz power frequency signal, the frequencies obtained by the formula (20) are sequenced, and the frequency obtained by subtracting the power frequency from other frequencies except the minimum frequency is the frequency of the voltage flicker component;
in terms of amplitude estimation, the expression of the source signal in the complex-valued independent component analysis model can be found from equation (12):
si(t)＝Piexp(jvit) (21)
in the formula (I), the compound is shown in the specification,
D0＝A0，A0is the amplitude of the power frequency signal, AiThat is, the amplitude of the voltage flicker signal component, and thus the amplitude of the flicker signal is obtained by taking the modulus value of the estimated signal obtained by analysis of the complex-valued independent component, i.e., the amplitude of the voltage flicker signal component
Di＝|si(t)|＝|Piexp(jvit)| (22)
Because the amplitude of the power frequency signal is far greater than that of the flicker signal component, D obtained by the formula (22) is obtainediSorting is carried out, the maximum value is the amplitude of the power frequency signal, DiAnd multiplying the other values except the maximum value by 2 times after sorting to obtain the amplitude of the voltage flicker signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810969807.4A CN109100562B (en) | 2018-08-24 | 2018-08-24 | Voltage flicker parameter detection method based on complex value independent component analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810969807.4A CN109100562B (en) | 2018-08-24 | 2018-08-24 | Voltage flicker parameter detection method based on complex value independent component analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109100562A CN109100562A (en) | 2018-12-28 |
| CN109100562B true CN109100562B (en) | 2020-06-02 |
Family
ID=64851172
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810969807.4A Active CN109100562B (en) | 2018-08-24 | 2018-08-24 | Voltage flicker parameter detection method based on complex value independent component analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109100562B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111308234A (en) * | 2019-12-09 | 2020-06-19 | 国网青海省电力公司 | Method for extracting S-transform power quality disturbance characteristics of Blackman window and window width ratio |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103116064A (en) * | 2013-02-06 | 2013-05-22 | 湖南大学 | Method and device for detecting voltage fluctuation and flicker based on energy operator and spectrum correction |
| CN103471848A (en) * | 2013-08-20 | 2013-12-25 | 哈尔滨工程大学 | Rolling bearing fault feature extraction method based on independent component analysis and cepstrum theory |
| CN103743949A (en) * | 2014-01-06 | 2014-04-23 | 国家电网公司 | Detection method of harmonic and inter-harmonic based on single-channel FastICA (Fast Independent Component Analysis) |
| CN104714075A (en) * | 2015-02-05 | 2015-06-17 | 湖南大学 | Method for extracting power grid voltage flicker envelope parameters |
| CN105282067A (en) * | 2015-09-16 | 2016-01-27 | 长安大学 | Complex field blind source separation method |
| CN205091399U (en) * | 2015-06-25 | 2016-03-16 | 云南电网有限责任公司电力科学研究院 | Voltage fluctuation and flickering source location detection device |
| CN105845148A (en) * | 2016-03-16 | 2016-08-10 | 重庆邮电大学 | Convolution blind source separation method based on frequency point correction |
| CN106569979A (en) * | 2015-10-13 | 2017-04-19 | 国网河南省电力公司信阳供电公司 | Voltage fluctuation and flicker signal detection algorithm |
| WO2017069727A1 (en) * | 2015-10-19 | 2017-04-27 | Halliburton Energy Services, Inc. | T1 distribution-based logging systems and methods using blind source separation independent component analysis |
| CN107085134A (en) * | 2017-04-01 | 2017-08-22 | 北京机械设备研究所 | A kind of detection method of voltage flicker |
| CN107390025A (en) * | 2017-07-31 | 2017-11-24 | 集美大学 | Power system method for distinguishing multiple harmonic sources based on blind source separating |
| CN108169540A (en) * | 2017-12-13 | 2018-06-15 | 内蒙古工业大学 | A kind of measuring method of wind power generating set voltage flicker |
-
2018
- 2018-08-24 CN CN201810969807.4A patent/CN109100562B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103116064A (en) * | 2013-02-06 | 2013-05-22 | 湖南大学 | Method and device for detecting voltage fluctuation and flicker based on energy operator and spectrum correction |
| CN103471848A (en) * | 2013-08-20 | 2013-12-25 | 哈尔滨工程大学 | Rolling bearing fault feature extraction method based on independent component analysis and cepstrum theory |
| CN103743949A (en) * | 2014-01-06 | 2014-04-23 | 国家电网公司 | Detection method of harmonic and inter-harmonic based on single-channel FastICA (Fast Independent Component Analysis) |
| CN104714075A (en) * | 2015-02-05 | 2015-06-17 | 湖南大学 | Method for extracting power grid voltage flicker envelope parameters |
| CN205091399U (en) * | 2015-06-25 | 2016-03-16 | 云南电网有限责任公司电力科学研究院 | Voltage fluctuation and flickering source location detection device |
| CN105282067A (en) * | 2015-09-16 | 2016-01-27 | 长安大学 | Complex field blind source separation method |
| CN106569979A (en) * | 2015-10-13 | 2017-04-19 | 国网河南省电力公司信阳供电公司 | Voltage fluctuation and flicker signal detection algorithm |
| WO2017069727A1 (en) * | 2015-10-19 | 2017-04-27 | Halliburton Energy Services, Inc. | T1 distribution-based logging systems and methods using blind source separation independent component analysis |
| CN105845148A (en) * | 2016-03-16 | 2016-08-10 | 重庆邮电大学 | Convolution blind source separation method based on frequency point correction |
| CN107085134A (en) * | 2017-04-01 | 2017-08-22 | 北京机械设备研究所 | A kind of detection method of voltage flicker |
| CN107390025A (en) * | 2017-07-31 | 2017-11-24 | 集美大学 | Power system method for distinguishing multiple harmonic sources based on blind source separating |
| CN108169540A (en) * | 2017-12-13 | 2018-06-15 | 内蒙古工业大学 | A kind of measuring method of wind power generating set voltage flicker |
Non-Patent Citations (4)
| Title |
|---|
| Interharmonic estimation based on complex valued independent component analysis;Zhao Liquan;《2010 IEEE International Conference on Wireless Communications,Networking and Information Security》;20100805;全文 * |
| 基于改进的FastICA算法的电压闪变检测;汪斌;《电力系统保护与控制》;20110901;第39卷(第17期);全文 * |
| 基于盲源分离理论的闪变和间谐波检测技术研究;国添栋;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20120315;全文 * |
| 独立分量分析及在信号提取中的应用研究;贾雁飞;《中国博士学位论文全文数据库 信息科技辑》;20190115;第5章第3节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109100562A (en) | 2018-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108469560B (en) | Electromagnetic interference objective complexity evaluation method based on rapid S-transform time-frequency space model | |
| CN107832777B (en) | Electric energy quality disturbance identification method adopting time domain compression multi-resolution rapid S-transform feature extraction | |
| CN110186682B (en) | Rolling bearing fault diagnosis method based on fractional order variation modal decomposition | |
| Tasitsiomi et al. | Modeling galaxy-mass correlations in dissipationless simulations | |
| CN106920232B (en) | Gradient similarity graph image quality evaluation method and system based on conspicuousness detection | |
| Das et al. | Diagnosis of power quality events based on detrended fluctuation analysis | |
| CN103279957B (en) | A kind of remote sensing images area-of-interest exacting method based on multi-scale feature fusion | |
| CN109255358B (en) | 3D image quality evaluation method based on visual saliency and depth map | |
| CN107506330A (en) | A kind of variation mode decomposition algorithm parameter optimization method based on particle cluster algorithm | |
| CN108414833B (en) | Accurate estimation method for signal component frequency | |
| CN104793124B (en) | On-off circuit method for diagnosing faults based on wavelet transformation and ICA feature extractions | |
| CN110068759A (en) | A kind of fault type preparation method and device | |
| CN109214469B (en) | Multi-source signal separation method based on non-negative tensor decomposition | |
| CN110110738A (en) | A kind of Recognition Method of Radar Emitters based on multi-feature fusion | |
| CN109034042B (en) | Non-stationary signal processing method based on generalized linear frequency modulation double-synchronous extraction transformation | |
| CN110006652A (en) | A kind of Fault Diagnosis of Roller Bearings and system | |
| CN109785272A (en) | A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction | |
| CN109100562B (en) | Voltage flicker parameter detection method based on complex value independent component analysis | |
| CN105510706A (en) | High-precision undersampling frequency measurement method | |
| Shevgunov et al. | Averaged absolute spectral correlation density estimator | |
| Cukierman et al. | Magnetic misalignment of interstellar dust filaments | |
| Jia et al. | Convolutive blind source separation for communication signals based on the sliding Z-transform | |
| Jia et al. | Multiwindow nonharmonic analysis method for gravitational waves | |
| CN108152584A (en) | A kind of high ferro tractive power supply system harmonic wave Multi-path synchronous rapid detection method | |
| US20120121038A1 (en) | Processing Method And Device For Simulating And Adding Noise To Digital Signals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |